Objective To evaluate whether a pragmatic corrected fortification (CF) model achieves recommended target protein and calorie content of human milk (HM) for preterm infants when compared with standard fixed-dose fortification (SF).

Design In this prospective non-interventional study, we enrolled mothers of infants with birth weight ≤1500 g fed exclusive HM. Infants with chromosomal or intestinal disorders were excluded. A total of 405 HM samples from 29 mothers and 45 donor milk samples were analysed for macronutrient content using a real-time HM analyser. A stepwise CF model was derived based on published data on HM calorie and protein content corrected for lactation stage and milk type. We applied both models to the measured protein and calorie content for all HM samples and compared the proportion of samples achieving target nutrient requirement in each group.

Results Target protein and calorie content of feed was achieved in 68% of HM samples with CF, compared with 5% samples with SF model (p<0.0001). For mother’s own milk, none of the samples met the target macronutrient range with SF fortification during later lactation periods (≥week 5). With SF, over 40% of infants had poor growth (decline in weight z-score ≥0.8 SD) by 8 weeks. The final feed osmolality was acceptable for all fortification steps of the CF model.

Conclusion The proposed CF model significantly improved the final protein and calorie content of HM with acceptable osmolality. It provides a proactive option to improve nutrient intake in premature infants.